Eighteen studies (n= 3,834) were identified: 14 randomised controlled trials (RCTs); three pre- and post-test studies; and one crossover trial with pre- and post-test design. The sample sizes of the trials ranged between 35 and 629. Follow up ranged between three and 12 months (average eight months). The quality of three trials was rated good; the other 15 studies were rated fair. The mean age of the populations included was 60 years. Attrition rates were higher than 15% in eight studies.
Physical functioning: Six of 12 studies evaluated this outcome and reported statistically significant improvements in physical functioning in studies that enrolled patients with arthritis, diabetes, congestive obstructive pulmonary disease (COPD) and "all conditions". Three studies of patients with arthritis reported improved joint function and protection. Significant improvements were found in exercise undertaken in two of three studies that evaluated this outcome.
One study of five that evaluated pain management reported a modest improvement. Medication adherence was found to be improved in only one study that assessed management of pain.
Illness knowledge: Three of six studies reported increased knowledge by patients about their conditions; two studies of diabetic patients observed that the increase in illness knowledge was associated with subsequent beneficial behaviour change.
Psychological functioning: Three of the six studies of patients with arthritis found significant decreases in depression and anxiety. Seven studies evaluated changes in self-efficacy, which increased after the intervention in five studies. Quality of life scores were observed to be significantly higher in two out of seven studies.
Health service use: Seven studies evaluated service usage and three reported mixed results.